KR102345753B1 - Method for intelligently visualizing data using a plurality of different artificial neural networks - Google Patents

Abstract

An embodiment of the present invention provides an intelligent data visualization system using a plurality of different artificial neural networks implemented in a computer system.

Description

How to intelligently visualize data using multiple different artificial neural networks

The present invention relates to a method for intelligently visualizing data using a plurality of different artificial neural networks.

In a situation in which the Internet has become common in countries around the world and the importance of data transmitted through it is increasing day by day, various technologies for more effective data utilization are rapidly developing. In particular, recently, data visualization technology that processes various data into various contents such as charts, images, and videos so that users can check various data at a glance has been in the spotlight.

This data visualization technique is largely composed of a technique for acquiring data to be visualized and a technique for visualizing the acquired data in a client environment.

Conventionally, in order to visualize a large amount of data, a user has to manually select a chart suitable for the data to be visualized, and perform tasks such as setting an axis for drawing the chart one by one. For this reason, the result of data visualization is highly dependent on the user's chart comprehension, and biased analysis may occur, and the effort and time required for data visualization are also greatly required, resulting in a decrease in productivity.

Korean Patent Publication No. 10-1910179 (2018.10.19.)

An object of the present invention is to provide a method for intelligently visualizing data using a plurality of different artificial neural networks.

An object of the present invention is to provide an intelligent data visualization system that performs high-accuracy data analysis without the user's expertise and automatically performs data visualization analysis based on data obtained regardless of the user's intention.

Another object of the present invention is to provide an intelligent data visualization system that provides predicted data using machine learning based on given data.

The problems of the present invention are not limited to the problems mentioned above, and other technical problems not mentioned will be clearly understood by those skilled in the art from the following description.

An intelligent data visualization system according to an embodiment of the present invention for solving the above problems is implemented in a computer system, a data table including a plurality of data each provided from a plurality of data sources, the plurality of data tables from the data table Machine learning that receives data and generates predictive data based on the data selected from the data acquisition unit using a data acquisition unit that receives data and selects at least one data from among the plurality of data, and a model previously learned through machine learning and a data visualization unit for generating data visualization contents based on the part and the selected data and the prediction data, wherein the pre-trained model includes at least one of a wide model and a deep model. It is a model generated by machine learning based on, and the wide model and the deep model each include an input node to which the selected data is input and an output node to which the prediction data is output. In case of category type data, the prediction data is generated using a model learned based on the wide model, and the wide model is a linear model including a first neural network in which the input node is directly connected to the output node. and, when the selected data is time series type data, the prediction data is generated using a model learned based on the deep model, and the deep model is at least one between the output node and the input node. a second neural network to which a hidden layer of The data visualization unit generates the data visualization content based on the visualization type recommendation information, and the visualization type recommendation information includes the selected data, the prediction data, and the image. It is derived based on variability information of pre-selected data and variability information between the selected data and prediction data, and the data visualization content includes a first visualization example and a second visualization example, and the first visualization example and the second visualization Examples include a content output unit that is displayed in the upper left and upper right order according to the recommendation order included in the visualization type recommendation information, and outputs a dashboard to the user, wherein the dashboard displays the data visualization content at least one panel, wherein the dashboard includes a main screen activated by a home button, a first sub screen activated by a first screen button, and a second sub screen activated by a second screen button, The main screen displays all visualization contents for the selected data, the main screen can remove some visualization contents among all the visualization contents according to a user's preset definition, and the first sub screen is the selected data from the selected data. When only the visualization content of categorical data is shown, the second sub screen shows only the visualization content of the time series data among the selected data, and the first sub screen and the second sub screen are activated at the same time, the dashboard is When divided, the first sub screen is displayed in a partial area, the second sub screen is displayed in the remaining partial area, and the main screen, the first sub screen, and the second sub screen are activated at the same time, the dashboard is When the main screen is displayed, the first sub screen and the second sub screen are displayed in the form of a pop-up screen, and one sub screen of the first sub screen and the second sub screen and the main screen are simultaneously activated; The dashboard represents the main screen, the one sub screen is shown in the form of a pop-up screen, and the visualization content is an image cone expressing data values as images. It includes a complex visualization type of visualization content including text content and text content expressing the data value as text, and the settings for the main screen, the first sub screen, and the second sub screen are recorded for each user, so that the user's provided in synchronization according to a login operation, and the content output unit enlarges or reduces the panel in response to a first user's input to the panel in the dashboard and displays the panel, and the first user's input is an edge of the panel , and the content output unit moves the panel in response to a second user's input to the panel in the dashboard, wherein the second user's input is a drag ( drag), and the content output unit displays a panel addition area in the dashboard, and provides a dashboard management Graphical User Interface (GUI) in response to a third user's input to the panel addition area, and the third user The input of includes at least one of a click and a double click on the additional panel, and when the panel addition area is activated according to the third user's input for the panel addition area, the The dashboard management GUI for the panel addition area appears in the form of a pop-up screen so as to overlap the main screen, and the dashboard management GUI includes a panel setting unit, and the panel setting unit includes a size, location, language and color of the additional panel. provides a set value for , wherein the first visualization example and the second visualization example each represent one of a line chart type, a bar chart type, a pie chart type, and a complex visualization type, and the type represented by the first visualization example and types indicated by the second visualization example may be different from each other.

Details of other embodiments are included in the detailed description and drawings.

According to the intelligent data visualization system according to the embodiments of the present invention, an intelligent data visualization system that performs high-accuracy data analysis without the user's expertise and automatically performs data visualization analysis based on the acquired data regardless of the user's intention can provide

In addition, according to various embodiments of the present invention, visualization content with high accuracy can be provided using machine learning based on a wide model and a deep model based on given data. , it can provide an intelligent data visualization system that provides predicted data to help users make quick decisions.

Effects according to the embodiments are not limited by the contents exemplified above, and more various effects are included in the present specification.

1 is a block diagram illustrating an intelligent data visualization system according to an embodiment of the present invention.
FIG. 2 is a block diagram illustrating a machine learning unit included in the intelligent data visualization system of FIG. 1 .
FIG. 3 is a diagram for explaining a first neural network included in the wide model of FIG. 2 .
FIG. 4 is a diagram for explaining a second neural network included in the deep model of FIG. 2 .
5 is an example of data visualization content output by the data visualization unit of FIG. 1 .
FIG. 6 is an example of a display screen in which a dashboard is implemented in a computing device by the content output unit of FIG. 1 .
7 to 9 are diagrams for explaining an interaction operation of the content output unit according to a user input to the dashboard.

Advantages and features of the present invention and methods of achieving them will become apparent with reference to the embodiments described below in detail in conjunction with the accompanying drawings. However, the present invention is not limited to the embodiments disclosed below, but will be implemented in a variety of different forms, and only these embodiments allow the disclosure of the present invention to be complete, and common knowledge in the technical field to which the present invention belongs It is provided to fully inform the possessor of the scope of the invention, and the present invention is only defined by the scope of the claims.

The shapes, sizes, proportions, angles, numbers, etc. disclosed in the drawings for explaining the embodiments are exemplary, and thus the present invention is not limited to the illustrated matters. Like reference numerals refer to like elements throughout. In addition, in the drawings, parts not related to the present invention may be omitted or simplified in order to clarify the description of the present invention.

Although the first, second, etc. are used to describe various elements, these elements are not limited by these terms, of course. These terms are only used to distinguish one component from another. Accordingly, it goes without saying that the first component mentioned below may be the second component within the spirit of the present invention. The singular expression includes the plural expression unless the context clearly dictates otherwise.

At this time, it will be understood that each block of the flowchart diagrams and combinations of the flowchart diagrams may be performed by computer program instructions of a computing device.

Additionally, each block may represent a module, segment, or portion of code that includes one or more executable instructions for executing specified logical functions. It should also be noted that in some embodiments it is also possible for the functions recited in blocks to occur out of order. For example, two blocks shown one after another may in fact be performed substantially simultaneously, or the blocks may sometimes be performed in the reverse order according to a corresponding function.

In this case, the term '~ unit' used in this embodiment means a hardware component such as software or an FPGA (Field-Programmable Gate Array) or ASIC (Application Specific Integrated Circuit), and '~ unit' means a specific can perform roles.

In describing embodiments of the present invention in detail, an example of a specific system will be mainly targeted, but the main subject matter to be claimed in the present specification is to extend the scope disclosed herein to other communication systems and services having a similar technical background. It can be applied within a range that does not deviate significantly, and this will be possible at the discretion of a person with technical knowledge skilled in the art.

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings.

1 is a block diagram illustrating an intelligent data visualization system according to an embodiment of the present invention.

Referring to FIG. 1 , the intelligent data visualization system 1000 includes a plurality of data sources 101 , 102 , 103 , a data table 200 , a data acquisition unit 300 , a machine learning unit 400 , and a data visualization unit. 500 , and a content output unit 600 .

The plurality of data sources 101 , 102 , 103 may include data retrieved from a database built on the Internet. Alternatively, the plurality of data sources 101 , 102 , 103 may be data uploaded from a database stored in the computing device. That is, it may be a data file uploaded by a user's simple selection operation.

The data table 200 may receive data DS1 , DS2 , DS3 from a plurality of data sources 101 , 102 , 103 and collect them, and may newly configure the database DATA according to the user's needs. can To this end, the first data source 101 provides the first data DS1 to the data table 200 , and the second data source 102 provides the second data DS2 to the data table 200 , , the third data source 103 may provide the third data DS3 to the data table 200 .

For example, each of the data DS1 , DS2 , and DS3 may be a spreadsheet method in which data is stored in units of cells. A variable type can be assigned to data stored in each cell. Variable types can represent properties of data. For example, each of the data (DS1, DS2, DS3) is classified into a predetermined specific type such as gender (male and female) and rank (private, private, corporal, sergeant), or by age, height, weight, etc. It can be divided into category type data classified as numerical values of , and time series type data having the same time series characteristics according to time sequence such as power consumption by hour and birth rate by year.

The data acquisition unit 300 may receive the database DATA from the data table 200 . In an embodiment, the data acquisition unit 300 may select at least one data from among the databases DATA of the data table 200 . The data acquisition unit 300 may automatically scan the database DATA to select data CDATA suitable for the user's needs. However, the present invention is not limited thereto, and the data acquisition unit 300 may select suitable data CDATA by scanning the database DATA according to a user-defined option.

The machine learning unit 400 may receive the data CDATA selected by the data acquisition unit 300 , and may generate the prediction data PDATA by learning the selected data CDATA as learning data. Here, the prediction data PDATA may be data derived by guessing information not provided in the selected data CDATA based on information given in the selected data CDATA.

According to an embodiment, the machine learning unit 400 may recommend a visualization type for effectively delivering information to a user in the process of learning the selected data CDATA. Such visualization type recommendation information may be provided to the data visualization unit 500 together with the prediction data PDATA.

The data visualization unit 500 may receive the prediction data PDATA and the selected data CDATA from the machine learning unit 400 to generate data visualization contents DVC. The data visualization content (DVC) may be data generated to deliver visual information to a user based on the data (CDATA) selected by the data acquisition unit 300 and the prediction data (PDATA) generated by the machine learning unit 400 . have. The format of the data visualization content (DVC) is not particularly limited as long as it is a format capable of effectively delivering data analysis content to the user, but may be in the form of a chart, image, video, text, or the like. Also, as described above, the data visualization unit 500 may determine the visualization type based on the visualization type estimation information provided from the machine learning unit 400 .

In an embodiment, the data visualization unit 500 may receive both the prediction data PDATA and the selected data CDATA from the machine learning unit 400 , but is not limited thereto, and the prediction data from the machine learning unit 400 is not limited thereto. The data PDATA may be provided and the data CDATA selected from the data acquisition unit 300 may be respectively provided.

The content output unit 600 may receive the data visualization content DVC generated by the data visualization unit 500 and display it to the user. The content output unit 600 may include a dashboard 700 for displaying data visualization content (DVC) and a control board capable of detailed editing of data visualization content (DVC) displayed in the dashboard 700, etc. have. The content output unit 600 may provide data analysis information to the user by displaying the provided data visualization content DVC on the dashboard 700 .

The content output unit 600 may perform an interaction corresponding to a user's input. A user's input may be made by hardware, including a mouse, camera, microphone, touch-driven screen, keyboard, and the like. The user's input may also be via other input mechanisms, such as optical gesture capture, gyroscope input devices, mouse, keyboard, eye tracking input, and comparable software and/or hardware based technologies. The user's input operation may be clicking (or double clicking) or dragging a partial area on the dashboard 700 .

As described above, according to the intelligent data visualization system 1000 according to the embodiment of the present invention, data visualization analysis can be automatically performed based on the acquired data regardless of the user's intention, so even without the user's expertise Data analysis with high accuracy can be performed.

In addition, the intelligent data visualization system 1000 according to an embodiment of the present invention can provide data visualization content (DVC) with high accuracy by using the given input data (CDATA) using machine learning, and predictive data ( PDATA) can be provided to help users make quick decisions.

FIG. 2 is a block diagram illustrating a machine learning unit included in the intelligent data visualization system of FIG. 1 .

1 and 2 , the machine learning unit 400 may include a wide model 410 , a deep model 420 , and a visualization recommendation unit 430 .

As described above, the machine learning unit 400 may perform machine learning using the data CDATA provided from the data acquisition unit 300 as learning data. Alternatively, the machine learning unit 400 may learn the learning data CDATA and the learning prediction data PDATA as learning data. Here, the machine learning unit 400 may apply at least one of the wide model 410 and the deep model 420 . That is, the machine learning unit 400 may learn the training data CDATA and the learning prediction data PDATA using the wide model 410 and/or the deep model 420 , and the input data CDATA and Prediction data (PDATA) can be output based on the model derived through learning. Here, both the training data CDATA and the training prediction data PDATA may be referred to as training data CDATA, and at least a portion of the training data CDATA may be used as the training data CDATA, and the remaining part may be predicted. It can also be used as data (PDATA). For example, when the training data CDATA is time series data, the last data or some data based on the temporal order may be used as the training prediction data PDATA, and the remaining data may be used as the training data CDATA. .

The machine learning unit 400 may perform machine learning by selecting a different model according to the type of the learning data CDATA. In one embodiment, the machine learning unit 400 may perform machine learning by applying the wide model 410 when the learning data CDATA is categorical data, and the learning data CDATA is time series. In the case of time series type data, machine learning may be performed by applying the deep model 420 . Also, the machine learning unit 400 may derive the prediction data PDATA based on a different model according to the type of the input data CDATA. In an embodiment, the machine learning unit 400 may derive the prediction data PDATA through the wide model 410 when the input data CDATA is category type data, and the input data ( When CDATA) is time series type data, prediction data PDATA may be derived through the deep model 420 .

However, the method by which the machine learning unit 400 selects a model is not limited thereto. According to an embodiment, the machine learning unit 400 may apply both the wide model 410 and the deep model 420 to perform machine learning on data, and derive predictive data for the input data by applying both. You may. Hereinafter, the wide model 410 and the deep model 420 will be described in more detail with further reference to FIGS. 3 and 4 .

The machine learning unit 400 may further include a visualization recommendation unit 430 . The visualization recommendation unit 430 may recommend a visualization type that can more effectively provide analysis information to the user based on the input data CDATA. In an embodiment, the visualization recommendation unit 430 may generate visualization type recommendation information based on the input data CDATA and provide it to the data visualization unit 500 . The data visualization unit 500 may determine and generate a type of data visualization content (DVC) based on the visualization type recommendation information provided from the visualization recommendation unit 430 .

FIG. 3 is a diagram for explaining a first neural network included in the wide model of FIG. 2 . FIG. 4 is a diagram for explaining a second neural network included in the deep model of FIG. 2 .

1 to 4 , the wide model 410 may include a first neural network 415 , and the deep model 420 may include a second neural network 425 .

The first neural network 415 and the second neural network 425 may include an input layer (IPL) and an output layer (OPL), respectively.

Data CDATA input to the machine learning unit 400 or data for learning CDATA may be provided to the input layer IPL. Various values included in the data CDATA may be respectively input to the input layer IPL.

A result value obtained by performing machine learning based on the learning data CDATA input to the input layer IPL may be output to the output layer OPL. According to an embodiment, learning may be performed by correcting a weight, etc. based on a result value according to the training data CDATA and the training prediction data PDATA. Also, the output layer OPL may output a result value obtained by machine learning based on the data CDATA input to the input layer IPL, that is, the prediction data PDATA.

The input layer IPL of the first neural network 415 may include a plurality of first input nodes 411 . The output layer OPL of the first neural network 415 may include a first output node 412 . In the first neural network 415 , the first input nodes 411 of the input layer IPL and the first output nodes 412 of the output layer OPL may be directly connected to each other. That is, another layer may not exist between the input layer IPL and the output layer OPL in the first neural network 415 . Accordingly, the wide model 410 may be a linear model including the first neural network 415 .

The input layer IPL of the second neural network 425 may include a plurality of second input nodes 421 . The output layer OPL of the second neural network 425 may include a second output node 422 . In the second neural network 425 , the second input nodes 421 of the input layer (IPL) and the second output node 422 of the second output layer (OPL) are not directly connected to each other, but may be indirectly connected to each other. have. That is, at least one hidden layer HDL may be connected between the input layer IPL and the output layer OPL to the second neural network 425 . Accordingly, the deep model 420 may be a nonlinear model including the second neural network 425 . Although only one hidden layer (HDL) is illustrated in FIG. 4 , the present invention is not limited thereto, and a larger number of hidden layers (HDL) may be connected between the input layer (IPL) and the output layer (OPL). The nodes of the hidden layer HDL may be respectively connected to the second input nodes 421 of the input layer IPL.

As described above, the intelligent data visualization system 1000 according to the present invention can classify the types of data to be analyzed into categorical data and time series data, and performs machine learning by applying different neural network models according to the data types. Therefore, it is possible to provide data visualization content with higher accuracy.

5 is an example of data visualization content output by the data visualization unit of FIG. 1 . In FIG. 5 , the data visualization content will be described as a chart including a bar graph and a line graph.

1 to 5 , the data visualization content DVC generated by the data visualization unit 500 may include an input area CR and a prediction area PR.

The input area CR may be an area expressing data given based on the selected data CDATA. The prediction region PR may be a region expressing the predicted value PV predicted by the machine learning unit 400 based on the selected data CDATA. For example, the data visualization content of FIG. 5 will be described as a visualization of time series data. As described above, when the selected data CDATA is time series data, the machine learning unit 400 may perform machine learning by applying the deep model 420 . Accordingly, the prediction region PR may indicate a value in a time period after the input region CR.

According to the data visualization system 1000 according to the present invention, the user can facilitate quick decision-making through the data visualization content DVC including the prediction data PDATA.

FIG. 6 is an example of a display screen in which a dashboard is implemented in a computing device by the content output unit of FIG. 1 .

1 and 6 , a dashboard 700 may be output by the content output unit 600 in the display area DA of the computing device. As an example, the content output unit 600 may refer to a client environment in the computing device for outputting the dashboard 700 .

In the present invention, the computing device includes a desktop PC, a tablet PC, a laptop PC, a communication device that can be connected to an external server through a network, such as IPTV including a set-top box, a mobile phone, a smart phone, a PDA (Personal Digital Assistant), All kinds of handheld-based wireless communication devices, multiprocessor systems, microprocessor-based or programmable consumer electronics, mini computers that can be connected to an external server through a network, such as PMP (Portable Multimedia Player), tablet PC, etc. , a mainframe computer, and the like.

The dashboard 700 may include a plurality of panels PN1 , PN2 , PN3 , and PN4 on which data visualization contents DVC1 , DVC2 , DVC3 , and DVC4 are respectively displayed. Here, the first panel PN1 displays the first data visualization content DVC1, the second panel PN2 displays the second data visualization content DVC2, and the third panel PN3 displays the third data visualization The content DVC3 may be displayed, and the fourth panel PN4 may display the fourth data visualization content DVC4.

As described above, visualization types that can be expressed by the intelligent data visualization system 1000 according to the present invention may be various. For example, the first data visualization content (DVC1) represents a line chart type, the second data visualization content (DVC2) represents a bar chart type, and the third data visualization content (DVC3) may represent a pie chart type.

Also, the fourth data visualization content DVC4 may be a complex visualization type including image content DVC4a expressing data values as images and text content DVC4b expressing data values as characters.

In the present embodiment, the data visualization contents DVC1 , DVC2 , DVC3 , and DVC4 may be changed according to a change in the input input data CDATA. For example, when a new value is added or changed in existing data, visualization contents may be updated based on the new data. The update of the visualization contents may be performed in real time, and in some embodiments, may be performed at a predetermined period according to a user's definition.

The dashboard 700 may include a home button 701 , a first screen button 702 , and a second screen button 703 .

The home button 701 may be a button that displays the main screen of the dashboard 700 . When the home button 701 is activated, the first screen button 702 and the second screen button 703 may be deactivated, and only the main screen may be displayed on the dashboard 700 . The main screen may be an area that comprehensively displays contents visualized by the data visualization unit 500 . The content output unit 600 comprehensively displays data visualization contents (DVC1, DVC2, DVC3, DVC4) on the main screen according to the activation of the home button 701, but some visualization contents are removed or added according to the user's definition can do.

For example, FIG. 6 illustrates a structure in which only the main screen is displayed on the dashboard 700 because the home button 701 is activated and the first screen button 702 and the second screen button 703 are deactivated.

The first screen button 702 and the second screen button 703 may be buttons for activating an area that can be defined according to a user's needs. For example, when the first screen button 702 is activated, a first sub screen may be displayed on the dashboard 700 , and only visualization contents of categorical data among data types are displayed on the first sub screen can do. In addition, when the second screen button 703 is activated, a second sub screen may be displayed on the dashboard 700, and only visualization contents of time series data among data types may be displayed on the second sub screen. have.

The first screen button 702 and the second screen button 703 may be respectively activated with each other, or may be activated simultaneously. When the first screen button 702 is activated and the second screen button 703 is deactivated, the dashboard 700 displays only the first sub screen, and vice versa, the dashboard 700 displays the second screen button 703 Only sub screens can be displayed. Also, when the first screen button 702 and the second screen button 703 are activated at the same time, the dashboard 700 may provide a divided screen. That is, the dashboard 700 may display the first sub screen in a partial area and display the second sub screen in the remaining partial area.

According to an embodiment, the home button 701 may be activated simultaneously with the first screen button 702 and/or the second screen button 703 . In this case, the main screen is displayed on the dashboard 700, and the first sub screen and/or the second sub screen according to the activation of the first screen button 702 and/or the second screen button 703 are pop- It may be displayed in the form of a pop-up screen.

In another embodiment, at least two screens among the main screen, the first sub screen, and the second sub screen may be provided as a split screen on the dashboard 700 .

For example, the home button 701 may be located in an upper region among the outer regions of the dashboard 700 , and a left boundary of the dashboard 700 and a left boundary of the home button 701 among the upper regions It can be positioned to be connected. In addition, the first screen button 702 may be located on the right side of the home button 701 , and the right boundary of the home button 701 and the left boundary of the first screen button 702 are separated by a predetermined distance. there may be In addition, the second screen button 703 may be located on the right side of the first screen button 702, and the right boundary of the first screen button 702 and the left boundary of the second screen button 703 are They can be separated by a certain distance. Here, the first distance between the home button 701 and the first screen button 702 and the second distance between the first screen button 702 and the second screen button 703 may be the same. However, since the first distance may be relatively farther than the second distance, the present invention is not limited thereto.

Screens displayed by the dashboard 700 and buttons therefor are not limited thereto. According to user definition, more sub screens and screen buttons can be created. These definitions may be recorded for each user and provided in synchronization according to a user login operation.

Meanwhile, the content output unit 600 may further display the panel addition area PN_ADD in the dashboard 700 . When the panel addition area PN_ADD is activated according to a user's input, the content output unit 600 may display a graphic user interface (GUI) for adding a new panel PN. A detailed panel PN addition method will be described later with reference to FIG. 9 .

7 to 9 are diagrams for explaining an interaction operation of the content output unit according to a user input to the dashboard.

First, referring to FIG. 7 , the size of the panel PN in the dashboard 700 may be freely adjusted according to a user input. For example, the user may drag the edge PNs of the panel PN using the pointer PT. The size of the panel PN may be adjusted according to a drag direction.

For example, when the edge PNs of the panel PN is dragged in the upper left direction in the drawing using the pointer PT, the size of the panel PN may be adjusted to the size of the reduced panel PNA. As another example, when the edge PNs of the panel PN is dragged in the lower right direction in the drawing using the pointer PT, the size of the panel PN may be adjusted to the size of the enlarged panel PNB.

Here, the user's input may be a drag operation with respect to the edges PNs of the panel PN, but is not limited thereto. For example, even when the user clicks or double-clicks the edges PNs of the panel PN, the size of the panel PN may be adjusted according to a preset definition. In addition, the size of the panel PN may be directly input and adjusted through a separate input command.

Next, referring to FIG. 8 , the location of the panel PN in the dashboard 700 may be freely moved according to a user input. For example, the user may drag the inner area of the panel PN using the pointer PT. A position of the panel PN may be moved according to a drag direction.

For example, when an inner region of the panel PN is dragged to the right in the drawing using the pointer PT, the panel PN may be moved to a position of the movable panel PNC.

Here, the user's input may be a drag operation on the inner region of the panel PN, but is not limited thereto. For example, the position of the panel PN may be directly input and moved through a separate input command.

Next, referring to FIG. 9 , the dashboard 700 may provide a panel addition area PN_ADD. When the panel addition area PN_ADD is activated according to a user's input, the dashboard 700 may further display a dashboard management graphical user interface (GUI) 710 . Here, the panel addition area PN_ADD may be activated by clicking or double clicking using the pointer PT.

The dashboard management GUI 710 output according to the activation of the panel addition area PN_ADD may be displayed to overlap the main screen of the dashboard 700 in the form of a pop-up screen, but is limited thereto. it is not going to be

The dashboard management GUI 710 may include a data list 711 , a panel setting unit 712 , and a visualization example unit 713 .

The data list 711 may display a list of data automatically scanned by the data acquisition unit 300 among the databases DATA included in the data table 200 . When one data among data of the data list 711 is selected according to a user's input, the selected data may be provided to the machine learning unit 400 as input data CDATA.

The panel setting unit 712 may specify information to be visualized among the selected data or provide setting values for setting a range to the user. The panel setting unit 712 may provide setting values for visualization content in the panel in addition to setting values related to data. For example, the panel setting unit 712 may provide setting values for the size, position, language, color, etc. of the newly added panel PN.

The visualization example unit 713 may display the first visualization example 713a and the second visualization example 713b to the user in response to the data selected from the data list 711 . The visualization examples 713a and 713b displayed by the visualization example unit 713 may be a visualization type for most effectively analyzing and providing selected data. As described above, the visualization example unit 713 may display visualization examples based on the visualization type recommendation information generated by the visualization recommendation unit 430 .

Meanwhile, when a user's definition of the visualization content of the panel PN exists by the panel setting unit 712 , the visualization example unit 713 may generate a visualization example according to the user's definition. Even in this case, the visualization example unit 713 may further display recommended visualization examples based on the visualization type recommendation information.

For example, the visualization type recommendation information may indicate at least one of a line chart type, a bar chart type, a pie chart type, and a complex visualization type. Alternatively, the visualization type recommendation information may be information in which the line chart type, bar chart type, and pie chart type are listed in a recommended order, and the visualization example unit 713 is based on the information listed in the recommended order, in the upper left corner , examples may be displayed in the order of upper right and lower left.

For example, the visualization type recommendation information includes a value of data CDATA, a value of predicted data PDATA, variability information of values of data CDATA, and variability between a value of data CDATA and a value of prediction data PDATA. It may be derived based on information, but is not limited thereto.

The first data visualization content DVC1 represents a line chart type, the second data visualization content DVC2 represents a bar chart type, and the third data visualization content DVC3 represents a pie chart ( Pie chart) type can be indicated.

Also, the fourth data visualization content DVC4 may be a complex visualization type including image content DVC4a expressing data values as images and text content DVC4b expressing data values as characters.

As described above, according to the intelligent data visualization system 1000 according to the present invention, the intelligent data visualization system 1000 provides the user with a plurality of panels PN that display various visualization contents on the dashboard 700 . A large amount of data can be effectively provided. In addition, since the data visualization content DVC provided through the panel PN may be changed in real time according to the data change, the user may react more immediately to the data change.

The type, size, location, etc. of the visualization content of the panel PN displayed on the dashboard 700 may be easily edited according to a user's input. That is, the size, movement, etc. of the panel PN can be immediately adjusted by the user's input on the dashboard 700 , and the user can use the panel addition area PN_ADD and the dashboard management GUI 710 on the dashboard 700 . ), you can directly check and add the visualization contents of the desired data, so user convenience can be further improved.

The embodiments of the present invention disclosed in the present specification and drawings are merely provided for specific examples in order to easily explain the technical contents of the present invention and help the understanding of the present invention, and are not intended to limit the scope of the present invention. That is, it will be apparent to those of ordinary skill in the art to which the present invention pertains that other modifications can be implemented based on the technical idea of the present invention. In addition, each of the above embodiments may be operated in combination with each other as needed. For example, all embodiments of the present invention may be implemented in parts in combination with each other.

In addition, the method for controlling the intelligent data visualization system according to the present invention may be implemented in the form of program instructions that can be executed through various computer means and recorded in a computer-readable medium.

As such, various embodiments of the present invention may be embodied as computer readable code on a computer readable recording medium in a particular aspect. A computer readable recording medium is any data storage device capable of storing data that can be read by a computer system. Examples of computer readable recording media include read only memory (ROM), random access memory (RAM), and compact disk-read only memory (CD-ROM). ), magnetic tapes, floppy disks, optical data storage devices, and carrier waves (such as data transmission over the Internet).

The computer readable recording medium may also be distributed over network coupled computer systems, so that the computer readable code is stored and executed in a distributed fashion. In addition, functional programs, codes, and code segments for achieving various embodiments of the present invention may be easily interpreted by programmers skilled in the field to which the present invention is applied.

In addition, it will be appreciated that the apparatus and method according to various embodiments of the present invention can be realized in the form of hardware, software, or a combination of hardware and software. Such software may include, for example, a volatile or non-volatile storage device, such as a ROM, or a memory, such as, for example, RAM, a memory chip, device or integrated circuit, whether erasable or rewritable, or For example, the storage medium may be stored in an optically or magnetically recordable storage medium such as a compact disk (CD), a DVD, a magnetic disk, or a magnetic tape, and a machine (eg, computer) readable storage medium. In addition, the software may be stored in a hardware module including a cloud server and/or a block chain server, or may be stored in a software module including a function of the cloud server and/or a block chain server.

In order to execute the method of chart visualization of a data table according to an embodiment of the present invention, access to the Internet environment may be made through a web browser. A web browser is a program that displays various information provided on the Internet through a web page, and may be installed together in the memory of each computing device when an operating system is installed, such as 'Internet Explorer', 'Chrome', 'Firefox' ', etc., may be separately installed in the memory of the computing device according to a customer request.

Accordingly, the present invention includes a program including code for implementing the apparatus or method described in the claims of the present specification, and a machine (computer, etc.) readable storage medium storing such a program. Also, such a program may be transmitted electronically through any medium such as a communication signal transmitted through a wired or wireless connection, and the present invention suitably includes the equivalent thereof.

Claims (1)

In an intelligent visualization system of data using a plurality of different artificial neural networks implemented in a computer system,
a data table including a plurality of data each provided from a plurality of data sources;
a data acquisition unit receiving the plurality of data from the data table and selecting at least one of the plurality of data;
a machine learning unit generating predictive data based on data selected from the data acquisition unit using a model learned in advance through machine learning; and
A data visualization unit for generating data visualization contents based on the selected data and the prediction data,
The pre-trained model is a model generated by machine learning based on at least one of a wide model and a deep model,
The wide model and the deep model each include an input node to which the selected data is input and an output node to which the prediction data is output,
The machine learning unit,
When the selected data is category type data, the prediction data is generated using a model learned based on the wide model, and the wide model is a first neural network in which the input node is directly connected to the output node. is a linear model including
When the selected data is time series type data, the prediction data is generated by using a model learned based on the deep model, and the deep model includes at least one hidden between the output node and the input node. A second neural network to which the layers are connected,
The machine learning unit further comprises a visualization recommendation unit,
The visualization recommendation unit generates visualization type recommendation information based on the selected data and provides it to the data visualization unit,
The data visualization unit generates the data visualization content based on the visualization type recommendation information,
The visualization type recommendation information is derived based on the selected data, the prediction data, variability information of the selected data, and variability information between the selected data and the prediction data,
The data visualization content includes a first visualization example and a second visualization example,
The first visualization example and the second visualization example are displayed in the order of the upper left and the upper right according to the recommendation order included in the visualization type recommendation information,
Further comprising a content output unit for outputting a dashboard to the user,
The dashboard includes at least one panel on which the data visualization content is displayed,
The dashboard includes a main screen activated by a home button, a first sub screen activated by a first screen button, and a second sub screen activated by a second screen button,
The main screen shows all visualization contents for the selected data, and the main screen can remove some visualization contents among all the visualization contents according to a user's preset definition,
The first sub-screen displays only the visualization content of the categorical data among the selected data,
The second sub-screen displays only the visualization content of the time series data among the selected data,
When the first sub screen and the second sub screen are activated at the same time, the dashboard is divided to display the first sub screen in a partial area and the second sub screen in the remaining partial area;
When the main screen, the first sub screen, and the second sub screen are simultaneously activated, the dashboard displays the main screen, and the first sub screen and the second sub screen form a pop-up screen,
When one of the first sub-screen and the second sub-screen and the main screen are activated at the same time, the dashboard displays the main screen, and the one sub-screen is displayed in the form of a pop-up screen,
The visualization content includes visualization content of a complex visualization type including image content expressing a data value as an image and text content expressing the data value as a character,
The settings for the main screen, the first sub screen, and the second sub screen are recorded for each user and provided in synchronization according to the user's login operation,
The content output unit displays an enlarged or reduced display of the panel in response to a first user's input to the panel in the dashboard;
The first user's input is a drag to the edge of the panel,
The content output unit moves the panel in response to a second user's input to the panel in the dashboard,
The second user's input is a drag to the inner area of the panel,
The content output unit displays a panel addition area in the dashboard, and provides a dashboard management Graphical User Interface (GUI) in response to a third user's input to the panel addition area,
The third user's input includes at least one of a click and a double click on the additional panel,
When the panel addition area is activated according to the third user's input to the panel addition area, the dashboard management GUI for the panel addition area appears in the form of a pop-up screen to overlap the main screen,
The dashboard management GUI includes a panel setting unit, and the panel setting unit provides setting values for the size, position, language and color of the additional panel,
The first visualization example and the second visualization example each represent one of a line chart type, a bar chart type, a pie chart type, and a complex visualization type, and the type represented by the first visualization example and the second visualization example represent An intelligent visualization system of data using multiple artificial neural networks of different types.

Images